Classifications

• • G—PHYSICS
  • G05—CONTROLLING; REGULATING
  • G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
  • G05B19/00—Programme-control systems
  • G05B19/02—Programme-control systems electric
  • G05B19/18—Numerical control [NC], i.e. automatically operating machines, in particular machine tools, e.g. in a manufacturing environment, so as to execute positioning, movement or co-ordinated operations by means of programme data in numerical form
  • G05B19/406—Numerical control [NC], i.e. automatically operating machines, in particular machine tools, e.g. in a manufacturing environment, so as to execute positioning, movement or co-ordinated operations by means of programme data in numerical form characterised by monitoring or safety
  • G05B19/4061—Avoiding collision or forbidden zones
• • B—PERFORMING OPERATIONS; TRANSPORTING
  • B25—HAND TOOLS; PORTABLE POWER-DRIVEN TOOLS; MANIPULATORS
  • B25J—MANIPULATORS; CHAMBERS PROVIDED WITH MANIPULATION DEVICES
  • B25J9/00—Programme-controlled manipulators
  • B25J9/16—Programme controls
  • B25J9/1674—Programme controls characterised by safety, monitoring, diagnostic
  • B25J9/1676—Avoiding collision or forbidden zones
• • G—PHYSICS
  • G05—CONTROLLING; REGULATING
  • G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
  • G05B2219/00—Program-control systems
  • G05B2219/30—Nc systems
  • G05B2219/39—Robotics, robotics to robotics hand
  • G05B2219/39084—Parts handling, during assembly
• • G—PHYSICS
  • G05—CONTROLLING; REGULATING
  • G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
  • G05B2219/00—Program-control systems
  • G05B2219/30—Nc systems
  • G05B2219/39—Robotics, robotics to robotics hand
  • G05B2219/39088—Inhibit movement in one axis if collision danger
• • G—PHYSICS
  • G05—CONTROLLING; REGULATING
  • G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
  • G05B2219/00—Program-control systems
  • G05B2219/30—Nc systems
  • G05B2219/39—Robotics, robotics to robotics hand
  • G05B2219/39091—Avoid collision with moving obstacles
• • G—PHYSICS
  • G05—CONTROLLING; REGULATING
  • G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
  • G05B2219/00—Program-control systems
  • G05B2219/30—Nc systems
  • G05B2219/39—Robotics, robotics to robotics hand
  • G05B2219/39097—Estimate own stop, brake time, then verify if in safe distance

Definitions

• the present invention relates to a method and a device for avoiding collisions between components of a multi-axial industrial robot and at least one other object.
• An object can be a physical part or an assembly of parts constituting a solid. Different types of objects may co-exist in the robot work cell.
• the object can either be moving or stationary. Moving objects are, for example, mechanical units such as robots, external axes of the robot, and positioners. Examples of stationary objects are rotation table, fixtures, and maintenance equipment. Collisions in a robot work cell may occur either between moving objects or between a moving object and a stationary object.
• US6 212 444 proposes to set a common area in which an operating area of a robot and an operating area of an apparatus, such as another multi-axial robot performing an operation cooperatively with the robot, overlap each other.
• a first entrance-forbidding signal is output to the apparatus forbidding the apparatus to enter the common area when a predeter- mined reference point on the robot is within the common area
• a second entrance forbidding signal is output to the robot forbidding the reference point of the robot to enter the common area when the movable part of the apparatus is within the common area.
• the other moving object is forbidden to enter the common area if a predetermined reference point, usually located at the tip of the mechanical arm of the robot or at a joint, is within the common area.
• a reference point on the robot is forbidden to enter the common area if the other moving object is within the common area.
• protective zones enclosing the moving objects or parts of the objects.
• the protective zones surround the mechanical arms of the robot and/or its tool. Protective zones are, for example, described in the follow- ing patent documents: US2005/0090930, JP2003-334777, and JP08-141978.
• US2005/0090930 describes a method where bounding boxes of various geometrical shapes are attached around parts of objects possibly to cover their whole bodies. Interferences between the boxes are checked by an algorithm, which judges whether the boxes overlap.
• protective envelops are used to model bodies of moving objects, whose positions are recorded together with their closest inter- distance at said positions, as well as size extensions of the envelops of the moving objects as functions of their speed. An in- terpolation of the recorded distances together with at least one predicted distance is accomplished and the minimum of the obtained interpolation curve is compared with the sum of the thicknesses of the envelops of the objects of concern.
• JP08- 141978 is described a method where a double protective en- velop is used around robot tools.
• US6678582 describes a method for collision avoidance between a manipulator and at least one other object.
• the method comprises: calculating a stopping time for the robot on the basis of actual and past joint positions of a multi-axial robot and velocities and accelerations of each robot joint, forecasting a configuration of a trajectory of the components of the robot at the stopping time, checking the predicted configuration through distance/interference algorithms for interference of the robot com- ponents with components of the other object, and stopping the robot and/or the other object in case a collision is imminent.
• the stopping time is the time it will take to stop the robot from the present state of the robot. Predicted configurations are determined at the estimated stopping time. It is assumed that the ro- bot continues to move with its present acceleration, from its present velocity and position.
• a shortcoming with the disclosed method is that large errors may arise in the predicted configurations, as they are only based on past and present joint positions, velocities and accelerations. For example, if the robot drastically changes its moving direction, there will be a large er- ror in the predicted configuration of the robot. Assuming that predicted configurations of several multi-axial robots at their respective stopping time are checked between each other, collisions between the robots could still occur since their predicted configurations are only based on past and present parameters. This brings about that large margins have to be used in order to avoid collisions between the robots.
• the object of the present invention is to provide an alternative to existing methods and devices for avoiding collisions, which provides a possibility to improve the prediction of future collisions and thereby reducing the margins necessary for avoiding collisions.
• this object is achieved by a method as defined in claim 1 .
• Such a method comprises repeatedly: estimating the time left until a collision occurs between the robot and the object, determining a stopping time for the robot, comparing the estimated time left until collision and the stopping time for the robot, and stopping the robot or the object in case the estimated time left until collision is close to the stopping time for the robot.
• the time left until a collision occurs between the robot and the object is estimated and compared with the stopping time of the robot, instead of, as in the prior art, comparing predicted configurations of the robot and the other object at the stopping time of the robot.
• the other object is either a stationary object or a moving object, such as another multi-axial robot.
• the distance between the robot and the ob- ject is increasing there is no imminent risk of a collision and the time left until collision will become infinite.
• An advantage of this approach is that a longer look-ahead time is achieved, which enables an earlier detection of a risk of collision and thus an improved estimation of when the robot should be stopped.
• a longer look-ahead time enhances the chance to detect imminent collisions and provides more time to decide what to do to prevent the collision and to stop the robot before the collision occurs.
• Another advantage achieved with the invention is that the margins necessary for avoiding collisions can be reduced.
• the method com- prises calculating the shortest distances between the components of the robot and the object at a plurality of future points in time and, on the basis thereof, estimating the time left until a collision occurs between the robot and the object.
• the method further comprises receiving information, such as information on future joint angles and information on a planned path for the robot. Said shortest distances between the components of the robot and the object are determined based on the planned robot path.
• a control system of an industrial robot includes a path planner, which plans the path of the robot.
• the path planner is adapted to receive movement instructions from an application program and, on the basis thereof, determine how the robot should move in order to execute the movement instruction.
• the path planner plans how the instructed movement should be performed by carrying out an interpolation of the movement.
• the interpolation includes dividing the instructed movement into a plurality of small increments, and computing the joint angles for all axes of the robot for each increment.
• the joint angles are then converted into motor references.
• the path planner trans- mits the computed motor references to drive modules of the robot.
• the path of the robot is planned a specified time interval ahead of the current position of the robot.
• information on the future path i.e. the actual planned path
• This information is already available in the path planner unit of the robot control system. Collisions are detected based on the programmed path.
• the actual planned path is used in order to predict whether a collision is imminent, instead of an estimated path calculated based on past movements, as in the prior art. This embodiment improves the accuracy of the estimation of whether a collision is imminent or not.
• the future points in time are consecutively selected during a defined look-ahead time interval.
• the look-ahead time interval depends on how far ahead the path is planned, i.e. the look-ahead time of the path planner, which depends on the control system of the robot.
• the method comprises performing an extrapolation of the stored shortest distances, until the distance between the robot and the object be- comes zero, and on the basis thereof determining the time left until a collision between the robot and the object.
• the shortest distances calculated based on the information from the path planner, form a shortest distance curve.
• the shortest distances curve is extrapolated using the calculated shortest distances at different future times.
• the extrapolated part of the curve is used to calculate the time left until collision.
• the look-ahead time then becomes the sum of the look-ahead time of the path planner and the extrapolated part of the curve. This embodiment further increases the look-ahead time and thus further enhances the chance to detect imminent collisions and provides more time to stop the robot before a collision occurs.
• the method comprises: defining one or more protective zones attached to and enclosing the components of the robot, defining one or more protective zones attached to and enclosing the object, calculating the shortest distance between the zone(s) of the robot and the zone(s) of the object at a plurality of future points in time, and based thereon estimating the time left until a collision occurs between the zones.
• a protective zone is defined as a vol- ume of an arbitrary or predefined geometry enclosing the whole or a part of the object/robot, which is to be protected from collisions.
• protective zones may surround the mechanical arms of the robot, the tool, or the workpiece carried by the robot.
• the protective zones are attached to the moving ob- ject, which means that they follow the movement of the object. This embodiment detects imminent collisions between predefined zones of the robot and the other object.
• the protective zones can be one, two or three-dimensional and have an arbitrary shape.
• the method comprises: defining at least two protective zones enclosing the components of the robot, calculating the shortest distance between the zones of the robot at a plurality of future points in time, and based thereon estimating the time left until a collision occurs between the zones of the robot.
• This embodiment enables detection of imminent collisions between predefined zones enclosing components of the robot. For example, this embodiment makes it possible to detect imminent collisions between a tool, or a work object carried by the robot and other parts such as arms of the robot. This embodiment is particularly advantageous if the robot carries a large work object, such as a car body side, and there is a risk that the work object comes into collision with the robot itself.
• said other object is a multi-axial industrial robot and the method comprises: defining at least two protective zones enclosing the components of the first multi-axial robot, defining at least two protective zones enclosing the components of the second multi-axial robot, forming pairs of zones consisting of one zone of the first robot and one zone of the second robot, calculating, for each pair of zones, the shortest distance between the zones at a plurality of future points in time, and storing the calculated shortest dis- tances in a buffer for each pair of zones, and based thereon estimating the time left until a collision occurs between the zones.
• This embodiment enables detection of imminent collisions between predefined zones enclosing different parts of two multi- axial industrial robots.
• the method further comprises calculating a shortest distance between the components of the robot and the object, determining whether the calculated shortest distance is smaller than a threshold value and in that case estimating the time left until collision occurs, otherwise no estimation of the time left until collision is done.
• a collision is only forthcoming if the distance between the robot and the object is decreasing, and as long as the distances between them are large enough there is no risk of an imminent collision.
• the time left until collision is only estimated if the shortest distance between the components of the robot and the object is shorter than a threshold value. Hence, needless computation is avoided and processor load is decreased.
• the robot or the object is stopped in case the estimated time left until collision is shorter than the stopping time multiplied by a safety factor set by the user.
• the robot and/or the ob- ject shall be ordered to stop in case the estimated time left until collision is close to the stopping time for the robot. If the robot is ordered to stop when the time left until collision is equal to the stopping time, there is a risk, due to calculation errors, that a collision will occur before the robot is stopped. Hence, to ensure that there is enough time to stop the robot, the time left until col- lision should be larger than the stopping time.
• the robot is ordered to stop in case the estimated time left until collision is shorter than the stopping time multiplied by a safety factor.
• the safety factor provides a safety margin for stopping the robot.
• This embodiment enables the user to set the safety factor and thus to determine the safety margin for stopping the robot.
• it can be useful to give the safety factor a value smaller than one for instance, when two robot trajectories presenting a low risk of collision have to be programmed in the vicinity of each other.
• the safety factor is set at a value below one, the collision avoidance detection for that pair of zones is turned off.
• the collision avoidance system can be temporarily turned off for a given pair of zones and then turned on again later with the help of robot command instructions.
• the method further comprises calculating a shortest distance between the components of the robot and the object, determining whether the cal- culated shortest distance is smaller than a defined minimum value and in case the shortest distance is smaller than the minimum value reducing the speed of the robot by a given percentage.
• a reduction of the speed of the robot results in a reduction of the stopping time of the robot, and thus the robot is permitted to come closer to the object until the collision avoidance system stops the robot.
• This embodiment also reduces the risk for false alarms from the collision avoidance system and thereby avoids unnecessary stops of the robot.
• the object is achieved by a computer program product directly loadable into the internal memory of a computer or a processor, comprising software code portions for performing the steps of the method according to the appended set of method claims, when the pro- gram is run on a computer.
• the computer program is provided either on a computer-readable medium or through a network.
• the object is achieved by a computer readable medium having a program re- corded thereon, when the program is to make a computer perform the steps of the method according to the appended set of method claims, and the program is run on the computer.
• this object is achieved by a device as defined in claim 14.
• Such a device comprises: a computation unit adapted to estimate the time left until a collision occurs between the robot and the object, to compute a stopping time for the robot, and to compare the estimated time left until collision and the stopping time for the robot, and to generate an order to stop the robot or the object in case the estimated time left until collision is close to the stopping time for the robot.
• Fig. 1 shows an example of a robot work cell including two multi-axial robots and a device for avoiding collisions according to an embodiment of the invention.
• Fig. 2 shows a planned robot path.
• Fig. 3 shows a shortest-distance curve for a pair of zones and illustrates how the time left until collision is calculated according to an embodiment of the invention.
• Fig. 4 shows flow diagram of a method for avoiding collisions according to an embodiment of the invention.
• a work cell including two multi-axial robots 1 ,2 and a stationary object 3, as shown in figure 1 .
• the collision avoidance device and method according to the invention can be used for any work cell including two or more objects, such as any combination of multi-axial robots, stationary or movable objects.
• a work cell may include one or more multi-axial robots.
• a work cell may include one or more stationary or moving objects.
• the robots 1 ,2 are connected to a common control unit 5.
• the control unit 5 is, for example, a conventional robot control system including necessary hardware, such as one or more proces- sor units, memory means, and input and output means, and software for carrying out normal robot control tasks.
• the control unit comprises program storage 7 for storing applications programs comprising program instructions written in a robot language, a program executor 8 adapted to execute the stored application programs, and a path planner unit 9 adapted to receive instructions from the program executor 8 and on the basis thereof determine how the robot or robots should move in order to execute movement instructions of the application program.
• control unit 5 includes a path planner entity 9 which can plan the path in synchronized or independent mo- tion for both robots 1 , 2 based respectively on separate application programs including linked movement instructions between both robots or separate application programs excluding such instructions for each robot.
• the control unit 5 also comprises a stop unit 10 adapted to stop the robot or the robots upon order.
• a collision avoidance unit 1 1 which monitors collisions between the robots 1 , 2 and the object 3 in the work cell, is located in the control unit 5.
• the collision avoidance unit 1 1 comprises a buffer unit 12 including a plurality of buffers for storing computed shorter distances between objects in the work cell, a computation unit 13 adapted to estimate the time left until a collision occurs between the objects and to compute stopping times for the robots, and to compare the estimated time left until collision and the stopping time for the robots.
• the collision avoidance unit 1 1 is adapted to send a stop order to the stop unit 10 in case the estimated time left until collision is close to the stopping time of the robot.
• the computation unit 13 monitors the robot 1 ,2 in order to detect whether a collision between the robot and the other objects in the work cell is imminent, and if a collision is imminent the computation unit generates an order to stop one of the robots, or both robots, which order is forwarded to the stop unit 10.
• the stop unit 10 receives the stop order from the computation unit 13, the robot or both the robots are braked until their movements are stopped.
• the robots are kept on their path during the braking.
• the collision avoidance unit 1 1 can be made as a separate hardware unit adapted to receive instructions from the path planner units of the robots 1 , 2. This is particularly advan- tageous in a robot system where the robots are provided with individual control units with separate path planner units.
• the collision avoidance unit 1 1 receives information on the planned path from the separate path planner units and communicates the stop order(s) to the robot control unit(s) concerned in order to avoid a collision.
• the stop signal from the computer unit 13 is commu- nicated to the safety system.
• the collision avoidance unit 1 1 further comprises a memory 16 for storing data rendered by defining protective zones attached to and enclosing the objects in the work cell.
• One or more pro- tective zones can be defined for each object.
• one protective zone 20 is defined for the stationary object 3 in figure 1
• two protective zones 21 , 22 are defined for the robot 1
• four protective zones 24 - 27 are defined for the robot 2.
• a protective zone may enclose one or more components of the robot, for example as for robot 2 in figure 1
• one protective zone may enclose each arm of the robot.
• a tool or a work object carried by the robot may be enclosed by its own protected zone. Zones are defined as volumes of arbitrary geometries.
• the volumes can range from simple geometric primitives, such as cylinders, spheres and boxes, to more general representations such as complex polyhedra and triangulated surface patches.
• Stationary objects and shared workspace areas may simply be enclosed in geometrical volumes.
• An object may be enclosed by a model of its whole body, a model of its surface, by virtual walls or by points only.
• simple CAD models of objects can be loaded into the robot control unit and be used as protective zones around objects.
• the computation unit 13 is adapted to calculate the shortest dis- tances between the protective zones of the objects of the work cell at a plurality of future points in time and on the basis thereof estimating the time left until a collision occurs between the protective zones in the work cell.
• the shortest distance is, for example, calculated by means of the well known Gilbert Johnsson Keerthi (GJK) algorithm.
• GJK Gilbert Johnsson Keerthi
• the shortest distances di-d 9 not all distances are shown in the figure
• the other protective zones in the work cell are calculated, possibly also the shortest distances between protective zones enclosing the same robot.
• robot 2 carries a large and extensive work object, surrounded by the protective zone 24, which may come into a collision with the other components of the robot, surrounded by the zones 25-27.
• the shortest distances d 5 -d 7 between zone 24, enclosing the work object, and the zones 25-27, enclosing the other components of the robot are calculated.
• the zones 21 , 22, 24 - 27 are movable zones, and the zone 20 is a stationary one. All possible pairs of zones in the work cell are determined. For example, if the work cell includes four protective zones, the number of possible pair of zones is six. In the example shown in figure 1 , the work cell includes seven protective zones and accordingly there are 21 possible pairs of zones.
• Each pair of zones is provided with its own buffer for storing the consecutive values of the shortest distances between the pair of zones. For each zone there exists a description file that contains information about the type, shape, dimensions, orientation, possible attachment of the zone to the object including the refer- ence coordinate system for the attachment of the zone to the object and a geometrical model of the zone.
• Figure 2 shows an example of a programmed robot path 30 to be followed by a robot.
• the arrow in the figure illustrates where the robot is positioned on the path at the present time.
• the path planner unity 9 calculates the future path of the robot by performing an interpolation of a movement instructed by the robot program. The instructed movement is divided into a plurality of small increments and the path planner computes joint angles for all axes of the robot for each increment.
• the figure illustrates past points P -4 - P -1 and future points P 1 - P 10 on the path.
• the path planner has planned the future points P 1 - P 10 in advance based on one or more program instructions in the robot program.
• the path planner has computed joint angles for ten future points.
• the time between each calculated future point is ⁇ t.
• the total time planned in advance by the path planner is called the look-ahead time t !a of the path planner.
• Figure 3 shows a diagram illustrating the shortest distance d sd between a pair of zones versus time t.
• the shortest distance between two zones is calculated in real time based on future joint angles from the path planner, a kinematic model of the robot, and geometric models of the protective zones.
• the configuration of the robot during the look-ahead time of the path planner can be determined based on the kinematic model of the robot and the joint angles from the path planner.
• the future configuration of the robot geometric models of the protective zones has been determined, it is possible to calculate positions on surfaces of the protective zones and on the basis thereof determining the distances between the protective zones.
• An iterative algorithm finds the shortest distance between the surfaces of the zones.
• the shortest distance between the zones is determined for each point P 1 -P 1 O from the path planner and is stored in the buffer unit 13.
• the positions of each pair of protective zones is predicted by calculating the positions of each moving object from a flow of joint angles supplied by at least one control unit for a given time of look ahead.
• the path planner calculates joint angles for the robot at ten future points.
• the buffer function has a time window and only contains a predetermined number of values.
• the buffer contains ten distance values for ten consecutive points in the time.
• the buffer may contain only fu- ture points in time P 1 - P 10 .
• Figure 3 shows the shortest distance values d sd stored in the buffer plotted as a function of time t. In this example, the shortest distance between the protective zones is decreasing.
• the stored distance values in the buffer are used to predict the time left until collision.
• the curve 32 is extrapolated until it intersects with the time axis. By then the distance between the zones is zero.
• the extrapolation is obtained, for example, by a linear or a 2 nd degree of interpolation curve.
• the point in time when the distance between the zones is zero is the estimated time left until collision t c .
• the time left until collision t c is equal to the sum of the look-ahead time of the path planner t !a and the time t e of extrapolated part of the curve 32.
• the shortest distances between each zone pair and the time left until collision are only calculated if the distance be- tween the protective zones is smaller than a predefined value d h , also denoted a horizon distance. As long as the distance between the protective zones is larger than the horizon distance, no computation of the time left until collision is made. A conservative approximation of the shortest distances is calculated with trees of bounding boxes around each protective zone. Thus, a horizon distance d h is defined, below which the shortest dis- tance between each pair of protective zones as well as the time left until collision will be estimated at regular time intervals.
• Data that must be communicated in real time to the collision avoidance unit are joint values. These are used as input to the kinematic models of the robots in order to calculate the positions of the components of the robot. The stopping times of the robots have to be calculated as well. These are obtained in real time based on a dynamic model for each robot and information contained in the path planner entity 9. Other data created and used in real time are distance computations stored for each zone pair in a buffer over a give time interval, an extrapolation function between the distances for each pair of zone in the time interval and a root of each of the extrapolation functions.
• Figure 4 is a flow chart illustration of a method and a computer program product according to an embodiment of the present in- vention. It shall be understood that each block of the flow chart can be implemented by computer program instructions.
• the algorithm described by the steps 44-64 is looped through for each pair of zones, until all pairs of zones have been taken care of, block 43.
• the shortest distance d sd between each pair of zones, or its approximation, is calculated, block 44, excluding stationary zone pairs.
• Each calculated shortest distance d sd is stored in a buffer belonging to the pair of zones, block 46.
• Each buffer already comprises a plurality of shortest distances calculated during previous iterations. The newly calculated distance overwrites the oldest calculated distance in the buffer.
• the calculated shortest distance d sd for the present pair of zones is compared with the horizon distance d h .
• the time left until collision tc for the pair of zones is determined based on the shortest distance values stored in the buffer and by means of an extrapolation operation, block 56.
• Information on axes positions are received from the path planner entity 9, block 58, and based thereon a stopping time t s for the robot is calcu- lated, block 60.
• the time left until collision is compared with the stopping time, block 62, and if the time left until collision is smaller than the stopping time multiplied by a parameter denoted a safety factor ⁇ , the robot or the object or both are ordered to stop, block 64.
• the safety factor ⁇ must be larger than one in order to provide a safety margin.
• the parameter ⁇ is defined such that at least one of the robot control systems shall order the robot to stop their movements if the predicted shortest time until collision is equal to or less than ⁇ times the stopping time of one of the robots involved in the impending collision.
• the robot operator is allowed to set or adjust the value of the safety factor ⁇ for each pair of zones.
• the operator is also allowed to set the safety factor ⁇ to a value less than one.
• the safety factor ⁇ is set to a value below one, the collision avoidance detection for that pair of zones is turned off. This is, for instance, advantageous if it is known to the operator that the path of the robot is close to, or coincides with, the object, and thus the robot should be allowed to come very close to, or even be in contact with, the ob- ject, without running the risk of being stopped by the collision avoidance system.
• the calculated shortest distance is compared with a defined minimum value d min , and in case the calculated shortest distance d sd is smaller than the minimum value, the speed of the robot is reduced by a given percentage, for example in the order of 20- 30% reduction of the speed. Reducing the speed of the robot means that the stop time t s of the robot is reduced and thereby increasing the allowed time left until collision t c according to equation 2.
• the speed of the moving objects to which the protective zones are attached and monitored shall be reduced by given percentage.
• the method comprises interpolating the 3 - 10 latest shortest distances recorded within the time interval by a linear or quadratic function. From the extrapolation of the obtained curve, estimate the time left until collision of pairs of objects. Send a command for moving objects to stop on path if the predicted time left until collision is less than ⁇ times their respective stopping time. Order an emergency stop for the robots, possibly off the programmed path.
• the predictions of the time left until collision are obtained with very high accuracy below a predefined threshold for the look-ahead time interval, whereas above the threshold they are obtained with lower accuracy.
• the predictions of the time left until collision are obtained with very high accuracy below a predefined threshold d h for the shortest distance between the pairs or zones, whereas above the threshold they are obtained with lower accuracy.
• the present invention is not limited to the embodiments disclosed but may be varied and modified within the scope of the following claims.
• the time left until collision as the time until the shortest distance curve intersects the time axis, i.e. the time when the distance between the robot and the object is zero

Landscapes

• Engineering & Computer Science (AREA)
• Human Computer Interaction (AREA)
• Manufacturing & Machinery (AREA)
• Physics & Mathematics (AREA)
• General Physics & Mathematics (AREA)
• Automation & Control Theory (AREA)
• Robotics (AREA)
• Mechanical Engineering (AREA)
• Manipulator (AREA)
• Numerical Control (AREA)
• Forklifts And Lifting Vehicles (AREA)
• Control Of Position, Course, Altitude, Or Attitude Of Moving Bodies (AREA)

Priority Applications (1)

Applications Claiming Priority (2)

Publications (1)

Family

ID=37728303

Family Applications (1)

Country Status (6)

Cited By (8)

Families Citing this family (63)

Citations (3)

Family Cites Families (2)

• 2006
  • 2006-09-14 AT AT06120615T patent/ATE412929T1/de not_active IP Right Cessation
  • 2006-09-14 ES ES06120615T patent/ES2316012T3/es active Active
  • 2006-09-14 EP EP06120615A patent/EP1901150B1/de active Active
  • 2006-09-14 DE DE602006003435T patent/DE602006003435D1/de active Active
• 2007
  • 2007-07-24 CN CN2007800332359A patent/CN101512453B/zh active Active
  • 2007-07-24 WO PCT/EP2007/057617 patent/WO2008031664A1/en active Application Filing

Patent Citations (3)

Non-Patent Citations (3)

Also Published As

Similar Documents

Legal Events